Summary

In this chapter, we focused on understanding the model explainability interface provided by H2O. First, we understood how the explainability interface provides different explainability features that help users get detailed information about the models trained. Then, we learned how to implement this functionality on models trained by H2O’s AutoML in both Python and R.

Once we were comfortable with its implementation, we started exploring and understanding the various explainability graphs displayed by the explainability interface’s output, starting with residual analysis. We observed how residual analysis helps highlight heteroscedasticity in the dataset and how it helps you identify if there is any missing information in your dataset.

Then, we explored variable importance and how it helps you identify important features in the dataset. Building on top of this, we learned how feature importance heatmaps can help you observe feature importance among all the...